Where Blended Finance Stops: Why Cheaper Capital Cannot Reach Organisations With No Surplus

Blended finance is usually described as a collection of bespoke transactions. It is not: beneath the instrument variety there is a single architecture, and this paper states it — five catalytic functions, separable from the instruments that implement them, arranged over a capital stack with an explicit time index. Stating the architecture makes its limit visible, which is the paper's main result and the reason for it. Every instrument in the family — concessional pricing, first-loss capital, guarantees, price support — works by making capital cheaper or safer. None of them changes whether an organisation has spare income to repay capital out of. An organisation can only take on capital, however cheap, if it generates a surplus to service it — and if the capital is made free, the structure has stopped being finance and become a grant. So there is no version of blended finance that reaches an organisation with no surplus: push the subsidy far enough and you arrive at philanthropy, which is outside the family. That boundary is where the population is. Read against Ghadamian's two-axis map, blended finance reaches the organisations conventional lenders already reach, plus a margin — and the organisations it was created for sit largely outside that, because cheaper capital changes the price of money and not whether there is income to repay it. Seven widely cited catalytic structures are re-read here and six fall inside the served quadrant, which is evidence for the boundary rather than for the architecture's reach. The conclusion is not that blended finance fails. It works, and it works exactly where it can — but where it can is fixed by the borrower's economics, not by how well the deal is structured, how large the fund is, or how carefully mobilisation is measured. Reaching past that line requires capital that takes no return at all, and the paper closes by saying plainly what that does and does not solve.